Thesis title
The impact of solar flares on Region D of the ionosphere : monitoring and modelling.
Composition of the jury
- Vincent Coudé du Foresto (Astronomer, LIRA) : Chair
- Elvira Astafyeva (Research Director, IPGP) : Rapporteur
- Jean-Pierre Raulin (Professor, Mackenzie Presbyterian University) : Rapporteur
- David Themens (Associate Professor, University of Birmingham) : Examiner
- Vincent Fabbro (Research Engineer, ONERA) : Examiner
- Angélique Woelfflé (Coordinator for the DGA) : Examiner
- Edmund Henley (Physicist at the Met Office) : Examiner
- Carine Briand (Astronomer, LIRA) : Thesis Supervisor
Abstract
Solar flares are eruptive events, accelerating particles and especially emitting energetic radiation. When reaching Earth, this additional source of ionisation increases the electron density in the ionosphere, and in particular in its lowest layer, the D-region (60–90 km). This causes enhanced absorption of High Frequency (HF, 3–30 MHz) waves, a frequency band regularly used for trans-horizon communications. Thus, continuous monitoring of the D-region is required, to alert in case of perturbation and mitigate the risks.
Satellites and balloons are not suitable for D-region monitoring due to its altitude range. Instead, Very Low Frequency (VLF, 3–30 kHz) waves propagating in the waveguide formed by the Earth and the ionosphere provide a continuous insight into the D-region’s electron density. Usually, VLF measurements are compared to the results from a propagation model, such as the Longwave Mode Propagator. By assuming that the ground conductivity is perfectly known, the electron density over the propagation path can be reconstructed by minimising the difference between the measurements and the modelling for various ionospheric conditions. However, we first showed that precise knowledge of the ground conductivity, in terms of both magnitude and spatial resolution, was crucial to accurately reconstruct D-region electron density profiles from VLF waves. This is particularly the case for low ground conductivity regions, like high latitude countries and mountain ranges.
The behaviour of the D-region is modelled by various chemistry schemes. We developed a new code, the Lower Ionospheric Region – Absorption and Chemistry Modelling (LIR-ACheM), to model the D-region. Since this model specifically includes ionisation by soft and hard X-rays, it is particularly suitable to describe the time evolution of the D-region during flares. LIR-ACheM considers seven main species in the D-region through the Mitra-Rowe scheme, offering a compromise between an accurate representation of the D-region and a low computation cost. The external forcing includes X-ray and EUV solar radiation. Through this model, the influence of the solar zenith angle, latitude and longitude, or seasonal variations may be investigated.
From LIR-ACheM, we showed that the delay of the electron density peak compared to the flux peak, and its recovery time, are controlled by the detachment at low altitudes and the ionisation by soft X-rays at high altitudes. A case study is also presented of two similar M1 flares occurring on the same day, causing different D region’s responses. We demonstrated that these differences were due to variations in the solar zenith angle and [NO] at high altitudes.
VLF waves are also very powerful real-time flare detectors. We developed a new tool, vlf4ions, from VLF data measured with a VLF AWESOME receiver in Nançay (Sologne, France). This tool enables flare detection from the changes in VLF phase slope, and the estimation of their flux by combining information from the VLF waves from several transmitters. It works independently from satellite data. Maps representing the D-region electron density on several propagation paths oriented in the North-South and Est-West direction are generated each minute, and alerts can then be sent to specific users as space weather services. This work thus represents an additional step towards a more resilient alert system for space weather events.
